We recently demonstrated that primitive hematopoietic progenitors expressing CD34, but lineage negative and HLA-DR (Lin34+DR) can give rise not only to myeloid progenitors in long-term bone marrow culture (LTBMC), but to natural killer cells (NK) of lymphoid origin in long-term natural killer cell culture systems (LTNK). We propose to examine in vitro the mechanisms associated with differentiation (Specific Aim 1) and maintenance/commitment (Specific Aim 2) of NK progenitors in the Lin34+DR - population, as well as maintenance of primitive progenitors within this population capable of both myeloid and NK differentiation (Specific Aim 3). In Specific Aim 1, mechanisms underlying NK progenitor differentiation will be explored. Preliminary data suggest that interrupting direct contact between marrow stroma and NK progenitors prevents NK progenitors differentiation in LTNK culture. We propose experiments employing marrow stroma, stromal cell lines and extracellular matrix components to identify the role of NK progenitor contact and correlation of CD3-zeta mRNA expression association with NK progenitor differentiation. These studies will identify specific receptor/ligand pairs implicated in NK progenitor differentiation in vitro. In Specific Aim 2, mechanisms underlying NK progenitor maintenance will be explored. Maintenance in vitro of NK progenitors in the Lin 34+DR- population also depend on contact with stroma; however, culture conditions differ from LTNK conditions required for progenitor differentiation. This suggests that distinct interactions with stroma dictate maintenance and differentiation of NK progenitors respectively. We will use a modified LTNK culture (NK maintenance culture) to identify mechanisms associated with direct contact with stroma which promote NK maintenance rather than differentiation. In Specific Aim 3 we will examine in vitro mechanisms which promote maintenance of single cells in the Lin34+DR- population capable of myeloid and of NK differentiation (multi-lineage progenitors). Maintenance of multi-lineage progenitors is optimal when Lin34+DR- cells are cultured in a stroma-non-contact system which separates progenitors from stroma by a microporous membrane. We hypothesize that maintenance of multi-lineage progenitors in this population is influenced by a balance of myeloid and lymphoid differentiating factors. We will culture the primitive progenitor population in stroma-non-contact culture and determine the effect of inducers (IL3, IL7) and inhibitors (MIP-1alpha, platelet factor 4 and AcSDKP) of differentiation on maintenance of a multi-lineage progenitor population. Single cell analysis and retroviral marking studies will be employed to document the common origin of myeloid and NK progenitors in sequential NK commitment and NK differentiation cultures. These studies will complement those of Project 1 characterizing the primitive human hematopoietic progenitor population. In addition, these studies will be essential to the development of strategies for ex vivo selection and expansion of primitive progenitor populations to be used in marrow transplant therapy of cancer as described in Projects 3, 4 and 5.